-
Notifications
You must be signed in to change notification settings - Fork 2
/
infer_HRVS.py
248 lines (205 loc) · 9.99 KB
/
infer_HRVS.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
from __future__ import print_function
import argparse
import os
import random
import torch
import torch.nn as nn
import torch.nn.parallel
import torch.backends.cudnn as cudnn
import torch.optim as optim
import torch.utils.data
from torch.autograd import Variable
import torch.nn.functional as F
import skimage
import skimage.io
import skimage.transform
import numpy as np
import time
import math
from utils import preprocess
from models import *
import shutil
import cv2
import glob
from utils import readpfm as rp
import matplotlib.pyplot as plt
from utils import init_spixel_grid
from utils.save_res import *
from utils.save_res import save_pfm
from utils.preprocess import get_transform
from skimage.transform import rescale, resize
#cudnn.benchmark = True
cudnn.benchmark = False
from PIL import Image
parser = argparse.ArgumentParser(description='PSMNet')
parser.add_argument('--KITTI', default='2015',
help='KITTI version')
parser.add_argument('--datapath', default='/home/fuy34/stereo_data/HR_VS/carla-highres/trainingF/',
help='select model')
parser.add_argument('--loadmodel', default=None,
help='loading model')
parser.add_argument('--model', default='prePSMNet_small',
help='select model')
parser.add_argument('--maxdisp', type=int, default=768,
help='maxium disparity')
parser.add_argument('--no-cuda', action='store_true', default=False,
help='enables CUDA training')
parser.add_argument('--seed', type=int, default=1, metavar='S',
help='random seed (default: 1)')
parser.add_argument('--savepath', metavar='DIR', default= '/home/fuy34/Dropbox/disp_res/HRVS_fixed_sp16' ,
help='save path')
parser.add_argument('--save_res', action='store_true', default=False,
help='save res')
parser.add_argument('--sz_list', type=float, default= [16], #, 8, 16
help='spixel loss weight')
parser.add_argument('--train_img_height', '-t_imgH', default=256*4, #384,
type=int, help='img height')
parser.add_argument('--train_img_width', '-t_imgW', default= 512*4, #768,
type=int, help='img width')
parser.add_argument('--val_img_height', '-v_imgH', default=2048, #512 368
type=int, help='img height_must be 16*n') #
parser.add_argument('--val_img_width', '-v_imgW', default=3072, #960 1232
type=int, help='img width must be 16*n')
parser.add_argument('--batchsize', type=int, default=4, #https://github.com/JiaRenChang/PSMNet/issues/73
help='number of epochs to train(default:12 or 8)')
parser.add_argument('--test_batchsize', type=int, default=4, #https://github.com/JiaRenChang/PSMNet/issues/73
help='number of epochs to train(default:12 or 8)')
args = parser.parse_args()
args.cuda = not args.no_cuda and torch.cuda.is_available()
torch.manual_seed(args.seed)
if args.cuda:
torch.cuda.manual_seed(args.seed)
# dataloader
from dataloader import listfiles as DA
_, _, _, test_left_img, test_right_img, disp_true = DA.dataloader(args.datapath, typ='trainval', b_HRonly=True)
# construct model
model = prePSMNet_small(args, None, b_finetune= True)
# model = spPSMNet(args)
model = nn.DataParallel(model, device_ids=[0])
model.cuda()
if args.loadmodel is not None:
pretrained_dict = torch.load(args.loadmodel)
model.load_state_dict(pretrained_dict['state_dict'])
else:
print('run with random init')
print('Number of model parameters: {}'.format(sum([p.data.nelement() for p in model.parameters()])))
def main():
processed = preprocess.get_transform(augment=False)
model.eval()
sum_epe = 0
for inx in range(len(test_left_img)):
# print(test_left_img[inx])
name = test_left_img[inx].split('/')[-2]
# if name not in ['exp-2_w-4_pos-10-61_00400'] : continue
if not os.path.isdir(args.savepath) and args.save_res:
os.makedirs(args.savepath)
imgL_o = Image.open(test_left_img[inx]).convert('RGB')
imgR_o = Image.open(test_right_img[inx]).convert('RGB')
# convert to quater resolution
w, h = imgL_o.size
# imgL_o = resize(imgL_o, (h // 4, w // 4), order=0)
# imgR_o = resize(imgR_o, (h // 4, w // 4), order=0)
imgL = processed(imgL_o).numpy()
imgR = processed(imgR_o).numpy()
imgL = np.reshape(imgL, [1, 3, imgL.shape[1], imgL.shape[2]])
imgR = np.reshape(imgR, [1, 3, imgR.shape[1], imgR.shape[2]])
# print(w, h, imgL.shape)
##fast pad
max_h = int(imgL.shape[2] // 64 * 64)
max_w = int(imgL.shape[3] // 64 * 64)
if max_h < imgL.shape[2]: max_h += 64
if max_w < imgL.shape[3]: max_w += 64
top_pad = max_h-imgL.shape[2]
left_pad = max_w-imgL.shape[3]
imgL = np.lib.pad(imgL,((0,0),(0,0),(top_pad,0),(0,left_pad)),mode='constant',constant_values=0)
imgR = np.lib.pad(imgR,((0,0),(0,0),(top_pad,0),(0,left_pad)),mode='constant',constant_values=0)
# test
imgL_in = Variable(torch.FloatTensor(imgL).cuda())
imgR_in = Variable(torch.FloatTensor(imgR).cuda())
with torch.no_grad():
torch.cuda.synchronize()
start_time = time.time()
pred_disp, maskL, _, _ = model(imgL_in, imgR_in)
torch.cuda.synchronize()
ttime = (time.time() - start_time) #print('time = %.2f' % (ttime*1000) )
pred_disp = torch.squeeze(pred_disp).data.cpu().numpy()
top_pad = max_h- h #imgL_o.shape[0]
left_pad = max_w- w #imgL_o.shape[1]
pred_disp = pred_disp[top_pad:,:pred_disp.shape[1]-left_pad]
# read gt
gt_disp = rp.readPFM(disp_true[inx])[0]
if args.save_res:
cv2.imwrite(args.savepath + '/{}_gt.png'.format(name), (gt_disp * 100).astype(np.uint16))
gt_disp[gt_disp == np.inf] = 0
gt_disp = np.ascontiguousarray(gt_disp, dtype=np.float32)
mask = (gt_disp != 0)
err = np.abs(gt_disp - pred_disp)*mask
epe = np.sum(err)/np.sum(mask)
sum_epe += epe
# # output
if args.save_res:
top_cut = 640
if not os.path.isdir(os.path.join(args.savepath, 'img')):
os.makedirs(os.path.join(args.savepath, 'img'))
img_save_path = os.path.join(args.savepath, 'img', name + '.png')
img = cv2.imread(test_left_img[inx])
print(img.shape)
cv2.imwrite(img_save_path, img[top_cut:])
MAX_DISP = 500
MIN_DISP = 0
# tgt_max = np.max(tgt_disp)
# tgt_min = np.min(tgt_disp)
if not os.path.isdir(os.path.join(args.savepath, 'tgt_disp_viz')):
os.makedirs(os.path.join(args.savepath, 'tgt_disp_viz'))
tgt_disp_save_name = os.path.join(args.savepath, 'tgt_disp_viz', name + '_gt.png')
plt.imshow(gt_disp[top_cut:], vmax=MAX_DISP, vmin=MIN_DISP) # val2uint8(tgt_disp, MAX_DISP, MIN_DISP)
plt.axis('off')
plt.gca().xaxis.set_major_locator(plt.NullLocator())
plt.gca().yaxis.set_major_locator(plt.NullLocator())
plt.subplots_adjust(top=1, bottom=0, left=0, right=1, hspace=0, wspace=0)
plt.margins(0, 0)
plt.savefig(tgt_disp_save_name, bbox_inches='tight', pad_inches=0)
# # save pred disp viz
if not os.path.isdir(os.path.join(args.savepath, 'pred_disp_viz')):
os.makedirs(os.path.join(args.savepath, 'pred_disp_viz'))
pred_disp_viz_save_name = os.path.join(args.savepath, 'pred_disp_viz', name + '_pred.png')
plt.imshow((pred_disp * mask)[top_cut:], vmax=MAX_DISP, vmin=MIN_DISP) # val2uint8(, MAX_DISP, MIN_DISP)
plt.axis('off')
plt.gca().xaxis.set_major_locator(plt.NullLocator())
plt.gca().yaxis.set_major_locator(plt.NullLocator())
plt.subplots_adjust(top=1, bottom=0, left=0, right=1, hspace=0, wspace=0)
plt.margins(0, 0)
plt.savefig(pred_disp_viz_save_name, bbox_inches='tight', pad_inches=0)
# err
if not os.path.isdir(os.path.join(args.savepath, 'disp_err')):
os.makedirs(os.path.join(args.savepath, 'disp_err'))
disp_save_path = os.path.join(args.savepath, 'disp_err', name + '_err.png')
skimage.io.imsave(disp_save_path, val2uint8(err[top_cut:], 20))
# spixel
img_l = torch.FloatTensor(imgL)
mean_values = torch.tensor([0.485, 0.456, 0.406], dtype=img_l.dtype).view(3, 1, 1).to(img_l.device)
std = torch.tensor([0.229, 0.224, 0.225], dtype=img_l.dtype).view(3, 1, 1).to(img_l.device)
img_l = (torch.FloatTensor(imgL) * std + mean_values).clamp(0,1).cpu().numpy()
_, _, h, w = imgL.shape
args.val_img_height , args.val_img_width = h, w
_, spixel_indx, _,_,_ = init_spixel_grid(args, args.sz_list, b_train=False)
maskL_viz, _ = update_spixl_map(spixel_indx, [maskL])
spixel_viz_L, _ = get_spixel_image(args, img_l[0].transpose(1, 2, 0), maskL_viz[0].squeeze())
print(spixel_viz_L.shape, img_l.shape, top_pad, left_pad)
spixel_viz_L = spixel_viz_L[:, top_pad:, :-left_pad ]
print(spixel_viz_L.shape)
if not os.path.isdir(args.savepath + '/spixel'):
os.makedirs(args.savepath + '/spixel')
dump_path = args.savepath + '/spixel/' + name + '_spixel.png'
skimage.io.imsave(dump_path, (spixel_viz_L.transpose(1, 2, 0))[top_cut:])
print('{}: {}'.format(name, epe))
print('meanEPE: {}'.format(sum_epe / len(test_left_img)))
torch.cuda.empty_cache()
def val2uint8(mat,maxVal, minVal=0):
maxVal_mat = np.ones(mat.shape) * maxVal
minVal_mat = np.ones(mat.shape) * minVal
mat_vis = np.where(mat > maxVal_mat, maxVal_mat, mat)
mat_vis = np.where(mat < minVal_mat, minVal_mat, mat_vis)
return (mat_vis * 255. / maxVal).astype(np.uint8)
if __name__ == '__main__':
main()